By Lisa Guerriero
SOASTA, Inc. is a performance measurement and analytics company with solutions for testing websites and Web applications. The company provides insight to help clients improve customer experiences. SOASTA reports having more than 10 million tests performed and over 100 billion user experiences measured.
Founded in 2006, SOASTA employs a staff of 250. The company is headquartered in Silicon Valley, CA, with offices worldwide. SOASTA is a privately held company and does not disclose its annual revenue.
Understanding the User Experience
In January, SOASTA announced the availability of its Data Science Workbench, a new query and analysis product that utilizes performance data to deliver insights from user experience. The company created the product from the ground up to take advantage of inexpensive cloud storage, as well as new query and reporting languages.
The Workbench builds off the existing SOASTA platform, using cloud-based storage to retain all available user experience data from its mPulse monitoring product. This, along with utilizing open source computing language, ensures fast transformation and loading of data.
“With SOASTA’s mPulse monitoring product, we realized that our customers were looking for a way to track customer user experience. Our solution collects information about the user experience, unlocking insight that could provide a competitive advantage to many of the customers we serve,” says Dan Bartow, principal product designer, SOASTA, of the solution’s origin.
The Workbench is designed to make the analysis of user experience data, both Web and mobile, simpler and more efficient. This includes understanding historic user conversion rates correlated to response time, viewing the most profitable click-paths by region, and observing peak user traffic by operating system. The user can then “slice, dice, and analyze” the information in a variety of visual formats, according to SOASTA.
The company emphasizes that data generated by activities like load testing and monitoring is valuable to ascertaining patterns and trends, particularly when integrated with other sources such as business and marketing. Their data is stored permanently, fostering insights over time as well as day to day.
The Need for Speed
SOASTA’s founders conceived the concept of the Data Science Workbench soon after the company acquired LogNormal in 2012, and development began in February 2014. Although SOASTA’s existing products were powered by real-time analytics engines, the company engineered the Workbench solution to offer something different.
“We collect terabytes of information for each of the top 10 retailers in the U.S. Whereas mPulse provides real-time access to this data, we wanted to give our customers a way to query their data ad-hoc and answer questions that might require additional compute time. We solve this big data problem through Data Science Workbench,” explains Bartow.
Data Science Workbench is available as an annual service package. It targets business owners, analysts, performance engineers, and data scientists that need to quickly gain insight, spot trends, or identify issues by mining user experience and performance data. It’s a good fit for organizations that need insight to improve customer experience, including all sectors and verticals, according to Bartow.
SOASTA believes clients need to arrive faster at the stage of evaluating the data they have access to. The company cites a 2014 New York Times article in which data scientists are estimated to spend from 50 to 80 percent of their time collecting and preparing data, before it can be analyzed. The Workbench is designed to help users avoid sifting through performance-related data and move on to extracting actionable insights. “Our solution provides a much faster path to data analysis than current available tools,” believes Bartow.
The product features pre-developed functions and statistical models to help users hone in on the most commonly asked questions. The company developed an interface, which is Julia-based, to enable users’ fast, streamlined exploration of large data sets. SOASTA’s Date Science team provides support to clients to ensure access to the information, as well as aiding with analysis if needed.
SOASTA partnered with 11 customers to develop the solution to provide data-backed answers to business queries. The Data Science Workbench is primarily intended for existing mPulse customers, but SOASTA also has clients interested in incorporating the solution into other data sources.
These customers are attracted to the speed and comprehensiveness of the solution, observes Bartow. “Data Science Workbench enables our customers to analyze their large data sets offline, and run large queries that are typically not done in a real-time system. We can scan years of data to find peak traffic timeframes, look at how performance has trended year over year, determine conversions, and help isolate issues that have led to site performance issues as well as most profitable click paths on a company’s site,” he explains.
Ready for Mining
There’s no doubt that the demand for information is growing, but that means data transformation and access must keep pace. With a strong background in performance testing, SOASTA harnesses its analytics experience and focuses it onto mining the user experience productively—and the Data Science Workbench is the result. SW